A surprising memory result deserves more than a headline. This article examines a specific historical case: Daryl Bem’s 2011 report and three attempts published in 2012 to reproduce one recall effect. It is not a review of every later study of precognition.
What was unusual about the sequence?
In Bem’s original paper, participants first saw words and completed a recall task. Only afterward did the computer select words for additional practice. Experiment 9 involved 50 Cornell undergraduates. Bem reported a difference favouring the later-practised words in the earlier recall task and interpreted it as a retroactive effect.
The important distinction is the timing: this was not simply a finding that ordinary practice before a test improves memory.
| Before later practice | After recall |
|---|---|
| Words were presented, then participants recalled them. | The computer selected a subset for practice. |
What happened in the 2012 replications?
Ritchie, Wiseman and French reported three preregistered attempts, involving 150 participants in total. None produced a significant effect supporting the retroactive-recall claim. Their procedure closely followed the original, using Bem’s software, with adjustments including British word forms and blinded coding of unrecognised responses.
These were close replications, not literally identical events. The paper provides the procedural details and limitations needed to judge the comparison.
Keep the conclusion at the right scale
The 2012 results did not reproduce this effect. That is a narrower conclusion than “every conceivable psychic claim has been disproved.” Conversely, the original positive report is not enough to certify someone’s ability or guarantee a future prediction.
A single study should not be converted into an advertisement that a particular reader knows what will happen to you. A research question, a statistical result and a paid service’s accuracy are different things.
For the distinction between an experience and an ability claim, see the psychic-gifts guide.
Five questions to ask about a dramatic study headline
- Which task was tested? A word-recall task is not automatically evidence about dreams, tarot, health or a workplace decision.
- What was recorded before the outcome? Identify the prediction and the result instead of relying on a story reconstructed later.
- Which comparison was made? Look for the actual conditions, not merely an impressive anecdote.
- What happened when another team tried it? Read the follow-up report, including relevant differences and limitations.
- What does the author conclude? Separate that interpretation from what you can responsibly infer for your own situation.
Do not turn a home game into a certificate
You may enjoy memory games or checking a written prediction against an outcome. But neither this article nor an informal score gives you a validated qualification to advise someone about health, money or legal matters.
Keep misses and unresolved results visible. Do not add meaning after the event until every answer appears correct, and do not blame a participant’s belief for an unfavourable result.
The creative imagery exercises deliberately distinguish memory and imagination from clairvoyance claims.
Read the claim and the follow-up together
Keep the original report, replication result and real-world decision separate. For reviewing a personal reading rather than an experiment, use the reading-review guide.
Frequently asked questions
1. What is this article’s scope?
A historical case involving a 2011 report and three 2012 replication attempts, not every later precognition study.
2. What did retroactive recall mean here?
The claim concerned earlier recall of words selected for practice later.
3. Was this ordinary practice before a test?
No. The additional practice followed the recall task.
4. How many participants were in Bem’s Experiment 9?
The original report describes 50 Cornell undergraduates.
5. How many people were in the three 2012 attempts?
There were 150 participants in total.
6. Did those attempts reproduce the effect?
None produced a significant effect supporting the retroactive-recall claim.
7. Were the procedures literally identical?
They were close replications with documented adjustments, not identical events.
8. Why read the procedural details?
They help you judge what was tested and how the comparison was made.
9. Does one positive report certify a reader?
No. A research result is not a qualification or an accuracy guarantee for a paid service.
10. Does this case settle every possible psychic claim?
The conclusion should remain tied to the particular effect and studies examined.
11. Can word recall establish tarot accuracy?
Do not transfer a finding from one task to an unrelated service.
12. What should a headline identify?
The task, prediction, comparison, result and relevant follow-up.
13. Is a home memory game a validated ability test?
Do not treat an informal score as certification.
14. Should misses remain in a record?
Yes. Keep misses and unresolved outcomes visible rather than rewriting them as matches.
15. How should I apply this to a reading?
Separate the interpretation from independently known information and do not use it as a professional decision guarantee.
It’s crucial to maintain a high standard of evidence in psychological research. The controversy caused by the original study emphasizes the need for transparency and replication to uphold credibility.
The debate around precognition and the scientific rigor in psychology is fascinating. It underscores the importance of replication studies to validate any extraordinary claims.
This case highlights the challenges within psychological research, especially when dealing with concepts that are difficult to quantify and replicate, such as precognition.
The controversy about the original study’s publication and the difficulties faced by opposing research reflects broader issues in the scientific community, particularly regarding publication bias.
The study presents an intriguing hypothesis, but the failure to replicate the results raises important questions about experimental design and the validity of the findings.